Isolation and contact tracing—which are now key topics as US officials discuss plans to open up the country—helped control the spread of COVID-19 in Shenzhen, China, according to a study published yesterday in The Lancet Infectious Diseases.
In the first known coronavirus research of its kind, researchers studied 391 COVID-19 patients and their 1,286 close contacts—identified through symptomatic surveillance and contact tracing from Jan 14 to Feb 12—to characterize disease course, transmission, and the effect of control measures.
After 622 of 653 close contacts with known quarantine dates were followed for at least 12 days, 98 tested positive, and one had presumed infection. Assuming that contacts with missing test results were not infected, the researchers estimated an attack rate of 11.2% (95% confidence interval [CI], 9.1 to 13.8) among household contacts and 6.6% (95% CI, 5.4 to 8.1) overall.
Risk of infection was highest for household contacts (odds ratio [OR], 6.27; 95% CI, 1.49 to 26.33) and those traveling with an infected person (OR, 7.06; 95% CI, 1.43 to 34.91). Infection was as common in children as in adults (7.4% in children younger than 10 years; average, 6.6% in overall population).
Mode of detection was known in 379 of the 391 patients. Of them, 292 (77%) were identified through symptom-based surveillance, and 87 were identified through contact tracing.
Rapid isolation of infected people
Patients identified through symptom-based surveillance were identified and isolated, on average, 4.6 days after symptom onset (95% CI, 4.1 to 5.0). Contact tracing reduced this time to 2.7 days (95% CI, 2.1 to 3.3).
The reproductive number (R0, or R-naught) was low, at 0.4 (95% CI, 0.3 to 0.5). The R0 indicates degree of infectiousness by showing how many people a single person can infect with a virus. If the R0 is less than 1, an outbreak dies; if the average is greater than one, it spreads.
The authors said that their analysis shows the effectiveness of contact tracing and isolation in reducing transmissibility of COVID-19, while noting that the overall impact of isolation and contact tracing is unclear and highly dependent on the number of asymptomatic patients.
"Contact-based surveillance in Shenzhen reduced the duration an infected individual transmits in the community by 2 days," they wrote. "We provide a key piece of evidence supporting intensive contact tracing and highlighting that children might be an important target for interventions aimed at reducing transmission, even if they do not get sick."
Mean serial interval, the time between successive cases, was 6.3 days (95% CI, 5.2 to 7.6). In the 183 patients with a well-defined exposure period and illness onset, the authors estimated a median incubation period of 4.8 days (95% CI, 4.2 to 5.4).
Infected patients were older than the rest of the population, with a mean age of 45 years, and were nearly equal in terms of sex. On initial assessment, 356 of 391 (91%) of patients had mild or moderate illness, while 35 (9%) were severe. While 330 (84%) of 391 patients had a fever, 25 (6%) were asymptomatic. In the contact-based surveillance group, 17 of 87 (20%) of patients had no symptoms at initial assessment.
As of Feb 22, three patients had died, and 225 had recovered, with a median time to recovery of 21 days (95% CI, 20 to 22).
In a commentary in the same journal, Kaiyuan Sun, PhD, and Cecile Viboud, PhD, of the US National Institutes of Health said that the serial interval in the study should be considered a lower bound that would likely increase in areas with less-controlled outbreaks.
Noting that contact tracing is labor-intensive and not always accurate when done manually, they called for new technology-based measures. "Building on the SARS-CoV-2 response in Shenzhen and other settings, we content that enhanced case finding and contact tracing should be part of the long-term response to this pandemic—this can get us most of the way towards control," they wrote.